as a Function
Knowledge and Culture
Cognitive Science Research
as a Function
Knowledge and Culture
1987 No. 18
Communications should be sent to:
Cognitive Science Research
S-223 50 Lund, Sweden
Coordinator: Bernhard Bierschenk
Department of Psychology
The study of consciousness requires a language for exact
and precise communication. This article presents a study of
how workers in various cultural contexts value information of
relevance to their job performance. The strategy of analysis
applied differs radically from the classical approaches. Based
on the concept of coordinative structure, functional linkages
among textual Agents and Objectives provide for the emergence
of mental phenomena. The experimental subjects were 35 randomly
selected mechanics from Sweden, England, West Germany, Italy,
and United States. Their verbal responses to three open ques- .
tions were analyzed. Only significant groupings that do not fall
below the lower limit of t nri of the t-distributions are re-
tained in the analysis. The subjective consciousness embedded
in the texts is visualized in a three-dimensional cubic space.
The results are discussed with respect to the cognitive process
depicted in the phase space. The experimental variable has given
rise to different mental structures. A comparison of what seems
to be prototypical of these cultural contexts shows a spectrum
from economic matters to basic questions of morality.
Within the social sciences, an ever growing number of stu-
dies have in some way or the other a need for obtaining verbal-
ized observations from their subjects. By this is meant that
here we are no longer concerned with relations among " observ-
ables " but with observations in Wigner's (1969) sense. The rea-
son for this change from collecting data obtained by rating scales
to the production of unrestricted text is a felt necessity to ac-
count for various cultural contexts and environments. Consequent-
ly, an unrestricted flow of ideas and the expression of opinions
in natural language is preferred to the processing of scores with
models of known processing capacity. The implicit assumption un-
derlying this approach may be stated by the following premise:
What humans observe as essential for becoming con- (1)
scious of their environment is accentuated by their
A representation of verbalized observations presupposes a perspec-
tive as well as the specification of a structure. Every verbal
description contains objective knowledge, but has to depict its
crucial structural qualities in such a way that intentionality
and orientation can be captured. When the purpose is to analyze
knowledge, it is central to differentiate between a subjective and
an objective component (De Mey, 1982). By applying the schema mo-
del, as outlined in Bierschenk (1984), it is assumed that the co-
operative actions taken by people in their cultural context re-
quire that the synthetic instead of the analytical proposition be
accounted for. This entails a second premise:
Natural language is the means to express an inten- (2)
tional and oriented schematization process
The discussion thus far may be summarized by the following formula:
Agent — +■ action — »► Objective (3)
Formula (3) incorporates the schema model, which imposes a con-
stitutive function on the context. Its assumptions are:
1 . The perspective of the agent governs the choice of view-
points, which means that the perspective is latent in the
2. The viewpoints are selected to describe objects and events
within a specified context.
3. The information embedded in the structure represented be-
comes accessible only in relation to activated components.
The kind of observation manifested in a verbal expression
depends on what is being realized through the action compo-
The generation of knowledge and its representation presuppose
models which manipulate the Objective component in the informa-
tion processing. The information gained consists of known facts,
that is, propositional knowledge and conditions. Information pro-
cessing based on a sequence of operations on propositional knowl-
edge, that is, conditionalized actions, may seem efficient or
rational, but can hardly extract or abstract the perspective
structure underlying the verbalization of an observation.
At times, the properties of the text produced indicate that
factual knowledge is sparsely distributed, a circumstance that
points towards its subordinate import. The function of such a text
is merely to emphasize the perspective of the producer. It is very
common that researchers in the social sciences have to deal with
this kind of text material, and even more common that they must
approach vast quantities with so called " unstructured " verbal
responses. This fact creates nearly un surmount able difficulties
when it comes to scientific analysis. These difficulties can be
overcome, but not unless a radically new position in designing a
strategy for text processing is taken.
The purpose of this article is to demonstrate the realizabi-
lity of a strategy that can
1. on the basis of a few coordinatively operating basic compo-
nents differentiate between viewpoints and perspective,
2. function as a basis for the construction of an algorithmic-
ally working system, and
3. identify an agent function as foundation for winning mean-
ingful conceptual relations.
This process always comprises on the one hand random language
elements with more or less frozen syntactic form, and constitu-
tive principles on the other. If chance shall be discriminable
from a principled order (Pribram, 1986), then a paradigmatically
founded theory is needed. This paradigmatic foundation as basis
for the strategy proposed rests on formula (3) and has been pre-
sented in Bierschenk (1982) and Bierschenk & Bierschenk (1976).
The methodological approach presented in Bierschenk 8c Bierschenk
(1986, a, b, c) has three distinctive advantages:
1. It is not bound to type of text or length of text.
2. It is applicable to at least European languages.
3. It is programmable.
The application of this strategy requires text processing within
a multivariate analysis system, if meaningful conceptual relations
shall be discovered.
In the following we will present the representation of struc-
tural relations in verbal observations from Swedish, English,
West German, Italian, and US workers of a multinational machine
industry. Based on data matrices suited for multivariable analysis,
cluster analyses have been carried out to represent verbal obser-
vations as a function of textual transformation and change. The
results will be graphically displayed and discussed with respect
to their conceptual structure. By keeping the perspective struc-
ture separated from the structure of the viewpoints, the method
used will demonstrate textual transformation in the form of a cu-
bic space. By means of the topographical representation, it will
be shown that text production is not only a divergent process,
but will be characterized by distinct cognitive phases, at times.
A Swedish multinational industry with companies in various
geographical areas of Europe and overseas has carried out an ex-
tensive survey. Their goal population amounts to about 30-40
thousand persons. The accessible population was estimated to
about 12-15 thousand mechanics. From this population, about 4-5
thousand individuals were selected by the representatives of the
companies. The material was divided up into the following cul-
tural contexts: Sweden, England, West Germany, Italy, and the
United States. Thus the cultural context functions as experimen-
tal variable in the analysis. It turned out that only 7 Swedish
subjects have produced useful unrestricted verbal responses. The
responses relate to three open-ended questions about information
management and use. Therefore, 7 subjects were randomly selected
from the other cultural contexts to match the Swedish sample.
The three questions posed to the workers have been transla-
ted by the industry. Because of variations in the wording, the
versions in the various languages will be given in extensor
Question 1 .
Swedish: Har du nagra ideer/f orslag pa hur man kan fa samtliga
mekaniker att anvanda servicehandbockerna?
English: Do you have any ideas/suggestions how to encourage more
mechanics to use the service manuals?
German: Haben Sie eigene Ideen/Vorschlage, wie man alle Machani-
ker zum Gebrauch der Service-Handbucher anregen kann?
Italian: Hai suggerimenti su come stimulare l'interesse alia let-
tura dei manuali di servizio, da parte dei meccanici?
Question 2 .
Swedish: Har du nagot f orslag pa hur informationen till dig kan
English: Do you have any suggestion how getting information to
you can be improved?
German: Haben Sie einen Vorschlag, wie man die Nachrichtenuber-
mittlung zu Ihnen verbessern kann?
Italian: Hai qualche proposta su come migliorare il sistema di
trasmissione delle informazioni al personale d'officina?
Question 3 .
Swedish: Tycker du att det sands ut for mycket papper (information)?
English: Do you think too much (or too little) paper (information)
is sent out?
German: Sind Sie der Meinung, dass zu viel Papier (Information)
Italian: Trovi che il materiale informativo inviato alle offi-
cine sia eccessivo?
These questions highlight different aspects of import to the
management and selective dissemination of information. To coun-
terbalance the variations in translation, they have been trea-
ted as if one question was posed. Some authentic responses will
now be reproduced to give the reader an impression of the kind
of text analyzed.
Swedish text . " Ta vack ackordlonesystemet sa att det
finns tid att lasa handbbcker = battre utfort jobb ". (Response
to question 1 . )
English text . " Far too much sent out. Judging by the
amount of bulletins sent out on the (A model] I would suspect
that it was not properly developed ". (Response to question 3.)
German text . " Bei uns im Betrieb werden die Mechaniker
immer auf gef ordert , regelmassig in die Handbticher hineinzu-
schauen. Es wird auch viel mit den Blichern gearbeitet ". (Res-
ponse to question 1 . )
Italian text . " Di farle piu semplificate perche da noi
in bassa Italia in media siamo pochi a riguardo la scuola almeno
per cid che riguarda della mia eta ". (Response to question 2.)
United States text . " I don't think too much paper infor-
mation can ever be provided. For most of the technicians this
is probably the only method by which they may stay informed " .
(Response to question 3.)
The attributes of the material are usually considered of
great importance for the definition of what is informative for
processing. Especially with artificially defined materials, expe-
rimentors are very careful in reporting the surface features, be-
cause these seem to be of import for their formulation and gene-
ration of knowledge. In our case, the surface features are compa-
ratively unimportant, because we are working with a naturally pro-
duced text, which is characterized by a structure existing beyond
what can be composed or computed by organizational connections.
Design and Procedure
The basic premise on which all scientific observation rest
is that processes can be isolated and studied with respect to
their structural stability. It was stated (3) that prototypical
forms are observable and that any consciousness that can be for-
malized into an expression such as
(A a ( A a 0, A a 0, . . . ) ) (4)
states that prototypical forms are recurrent. The formulas (3)
and (4) indicate transference, which means that the AaO's of the
text cooperate in order to lift the information from its syntac-
tic carriers. This requires the modelling of the AaO's such that
they can be formalized with respect to the following basic com-
Agent : denotes an action center, i.e., the point of refer-
ence for an observation. It is the intentional and
thus most basic component without which no action
can whether occur nor be controlled.
Action : denotes the intentional character of an observation
expressed. Thus the action is instrumental for the
Agent in coordinating the observations.
Objective : denotes the final point for the action and has the
following subcomponents :
Figure : denotes the point of intention.
Ground: denotes the point of orientation.
Means : denotes the instrument of an observation.
Setpoint : denotes the limit of an observation.
The procedural processing of running text will be exemplified
on the first sentence of the US text example. Table 1 illustrates
that every textual element gets its unique code assigned to it.
The first figure in the code addresses structural or dynamic pro-
perties, while the second one represents more static, i.e., de-
scriptive information. The output of this type of processing con-
stitutes the input for matrix construction. Every 30 code within
the boundaries of a 01 or 00 code makes up the rows of a matrix,
while all 50 codes constitute the columns of that matrix. Conse-
quently, every realization of a structural relation can be dis-
tinguished and counted. With respect to the material, matrices
have been built up of the Figure, and the Ground components for
each of the five cultural contexts separately. The relational
affinity between the Agent and the Objective components is de-
fined by the Action component, that is, every time a 40 code re-
lates an A to an 0, affinity is marked, otherwise independence.
A consequence for the matrix generation is that a not explicitly
stated Agent implies an X-variable, while an absent Objective
implies an Y-variable.
In order to specify the linkage in a group of textual Agents
whose activities covary as a result of shared Objectives, every
relation is assigned the binary code 1 independent of its fre-
quency. All other cells in the matrix are assigned a code. For
the reading in of the observed relations a protocol is set up
in which all relevant columns for a certain Agent are given. On
Table 1 . Formatting of text
Text Data Code Text Data Code
much 5 2
Note : The rules for processing and the scheme for formatting
are given in Bierschenk 8c Bierschenk (1986, b,c).
the basis of the protocol, a computer program builds up the ma-
trices as input for cluster analysis. This method is especially-
suited when we can assume affinity instead of the usual indepen-
dence. A variety of grouping methods of a data set can be execu-
ted according to various principles. The principles of relevance
are partly a successive agglomeration of elements, partly the
definition of an optimization criterion for governing the group-
ing process. The mechanics of the analysis permit the discovery
of coordinative structures of the Agents, which is a prerequi-
site for discovering the perspectives of the texts. Further,
the relations between the grouped Objectives of the texts can be
discovered, which means that we can state the coordinative struc-
ture of the viewpoints.
The first step in the analysis of the matrices requires the
choice of a clustering algorithm. Among the procedures available,
Ward's (1963) method seems to be sufficiently reliable for binary
data. He suggests the computation of the loss of information at
every step in an iteratively working grouping process. His opi-
nion is that loss of information can be measured by the total sum
of the squared deviations of the points (Agents, Objectives) from
the mean value of the cluster to which the respective point be-
longs. At every step in the analysis, the algorithm computes a
union of every possible pair of clusters and agglomerates those
two clusters whose combination results in a minimum loss of infor-
mation. The algorithm builds on an updating of a squared Euclidi-
an matrix of distances between cluster centroids. The solution
of the clustering results in a hierarchical organization with
quantitative estimations of a loss of information at every level
in the hierarchy. The minimum increase of the error sum of squares
is proportional to the squared distance between the centroids for
the agglomerated groups. This result differs from the centroid
method in that Ward's method implies a weighing of the distance
between the centroids when the distance is computed. The computa-
tional procedure has been executed with Wishart's (1982) CLUSTAN
When an analysis of the experimental texts takes its point
of departure in the schema model (formulas (3) and (4)), a strict
dependency between textual elements becomes visible. The model
may therefore be used for a transference and extraction of concep-
tual relations and their descriptions. The capacity of the schema
model to synthesize successive segments in a progressive devel-
opment of a text is founded on the steering and control function
of the Agent component. It controls perspective and change in
perspective at the same time as it indicates which intention is
operating on the textual level. In such a way it abstracts the
consciousness of the text producer. Presumably different Agents
have different functions in a text and have been chosen to give
expression to what is subjectively conscious. A differential ty-
pological analysis of the Agent function can give information
about which processes may be thought to operate in the manifes-
tation of the cultural influence on consciousness.
Any hierarchic grouping procedure produces dendrogram data.
These can be represented in the form of a tree. The presentation
of the tree as well as all other relevant information is usually
too extensive, however, to be reproduced in an article. Because
of this, we have chosen to give information about the groupings
that are significantly different together with the values of the
t-distributions and their confidence intervals. The rule used is
Rule 1: Upper Tail Rule, which is reported in the CLUSTAN manual
(Wishart, 1982, p 14-16). The way in which the test was applied
is described in Bierschenk & Bierschehk (1986 c, p 12). The re-
sults are given in Table 2. Only configurations building on signi-
ficant groupings which do not fall below the lower limit of the
t. qQ of the t-distributions are retained for topographic repre-
sentation and further analysis.
The cognitive process of the Swedish workers and the struc-
tural relations contained in the Swedish text are presented in
Figure 1 . The background of the cube takes up the Figure compo-
nent. At the base of the cube the relations pertaining to the
Ground component are visualized. The foreground represents the
perspective on the Figure. The description will begin with the
The edges of the planes represent the terminal states of
the process. The first terminal state named Labourousness in-
cludes most of the information in the Swedish text. This concept
is determined by the motivation of the workers, labour discipline,
rest hours, number of working days, and salary claims. In parti-
cular it addresses the workers' concern with getting paid for
the time units invested in their familiarization with service in-
formation. The next state initiating a short process is Techni-
Table 2 . Levels of significance for Agent, Figure, and Ground
components in Swedish material
Predicted Realized t =Vc x a and Percentiles
Clusters(c) Deviates(a) df of the t-distributions
Agent N = 15, n = 52, m = .135, s = .323
< t .95
= 15, m
s = .240
< t .975
< t .9C
< t .90
< t .95
< t .95
2, m =
Note : NS = non significant
The t-statistics with (C = cluster -l) number of degrees of free-
dom is computed by a multiplication of the deviations with the
square root of (C -l). vishart uses (n -i) where n = number of
values on the optimizing function. A C-based instead of an n-
based test value minimizes the problem of non-normality of the
Error Sum of Squares.
Unsophistication < r— <•.
Lack of Quality
Lack of Quality
Figure 1 . Operating structural relations
characterizing perspective and
viewpoints of Swedish mechanics.
From " A cognitive economics ap-
proach to information management
by B. Bierschenk, I. Bierschenk,
forsknina (13). p 1.
cal Simplification. This state gives expression to the demand
for simplified instructions, i.e., instructions that can go
into a pocket book. The process transits through the terminal
state named Simple-mindedness, which stands for the effect of
an expressed reluctance to invest efforts for reaching a certain
level of technical qualification. The transformation results in
the singularity of Unsophistication. This peak should be under-
stood as the absence of refinement or a structured approach to
instructional materials. This first step in the process is trans-
formed by Labourousness, which produces the second and final
singularity, Lack of Quality. The natural or essential character
of the work or goods produced makes it probable that they will
be of lesser standard than one would expect.
The perspective on the Figure extracts one singularity,
which is Lack of Quality.
There is no perspective on the Ground. The Ground itself
expresses a one-step process, which starts with the terminal state
Working Context. It refers to the conditions at the working place,
particularly substances polluting the environment, air condition-
ing, and cleanness. The second state, Updating of Message, re-
fers to the import of being continuously informed about ongoing
matters of value to the personnel. The transformation produces
Working Climate as the only singularity in the Ground.
Conclusions . The results presented point towards the po-
wer of natural language in carrying on the integration of expe-
rience into a mental structure. It is obvious that this structure
is quite rudimentary. The structural relations having emerged
through the numerical analysis and graphical presentation show
that Swedish workers are fairly unconcerned about information of
technical kind. Moreover, the Ground component brings out a con-
sciousness about ergonomic matters.
After the observations of the structural relationships char-
acteristic of Swedish mechanics in relation to the questions of
job related information, it would be of interest to study the
subjective consciousness of their English counterpart. The signi-
ficant groupings of the numerical analysis are presented in
Table 3. The table shows that six groupings have emerged,
which means that there is more structure in the English text com-
pared to the Swedish one. The cubic space containing this struc-
tural information is presented in Figure 2. The background of
the cube takes up the Figure component. At the base of the cube
the relations pertaining to the Ground component are visualized.
The foreground represents the perspective on the Figure. The de-
scription will begin with the Figure component.
The edges of the planes represent the terminal states of the
processes. The process of the Figure component starts in the
state of Model Criticism. It basically addresses the criteria for
the evaluation of the functional fitness of the Service Support
unit of the multinational industry. Such criteria could be
Table 3 . Levels of significance for Agent, Figure, and Ground
components in English material
Predicted Realized t =y~C x a and Percentiles
dusters(c) Deviates(a) df of the t-distributions
Agent N=12, n=49, m=.l78, s= .349
N = 49, n
= 12, m
s = .273
< t .975
< t .99
< t .975
N = 13, n
= 4, m =
< t .95
Organization |- ; :
"*< Infeasibility Information Quality
-;y \ Service Support
Information Structuring «/__
Figure P . Operating structural relations characterizing perspec-
tive and viewpoints of English mechanics.
(1) Significance: denoting the preciseness and validness of
the information with respect to the addressee's actual
(2) Exactitude: asking for information that describes an inno-
vation or modification reliably.
(3) Influence: requiring information that to a certain degree
has impact on a particular worker's motivation and actions.
This state becomes transformed by Lack of Information, which
addresses the extent to which disseminated information is struc-
tured to meet the information needs of a particular person or
group. The result of this transaction is the first singularity,
named Preparedness Restrictions. At this point, it becomes ob-
vious that the workers seem to experience a vague or ill-defined
conception of how to go about in finding a problem solution. It
may also mean that they are searching for too much information
or that they have acquired a habit of restricting their behaviour
in such a way that the problem is conceived to be unsolvable.
The perocess transits in a next step into the state of Em-
barrassment, meaning a state related to the valuation of what
can be performed and what cannot. This step produces Infeasibi-
lity. At this point a mental state is reached, which can be con-
ceived, as a conceptualized barrier towards carrying out service
functions. From here the process transits through the state of
Overabundance. This state indicates that required selection pro-
cesses do not function as expected. The resulting sungularity is
Discontentment. The stress at this point of the produced curve
is on a feeling of resentment towards the company.
The process now leads to a transit through the state of
Instructional Aid, which puts a heavy load on the personnel with
the task to give Service Support. Service Support is the singu-
larity which marks the import of determining the kind of informa-
tion needed in carrying out particular tasks for specified pur-
The final transition refers to the state of Organization of
Mediation and results in the highest point of the curve, called
Information Quality. Therefore it can finally be stated that
Service Support means the creation of information that necessi-
tates continually ongoing retransformation processes which are
prone to distortion.
Thus the root of the mental process depicted seems to con-
cern a questionable quality of the information disseminated.
This circumstance is further enhanced through the perspective on
the Figure. Its focus is on Information Quality.
The Ground component is defined by two terminal states, In-
formation Structuring and Information Access, leading to the sin-
gularity of Coherence. Information Structuring requires that the
personal concepts of the information user be used in information
management. Information Access refers to the time interval bet-
ween requests for information and its availability. Coherence is
the only singularity in the Ground and refers to the problem of
judging the relevance and meaningfulness of the information pro-
vided through the system's performance.
Conclusions . The English workers are from a topical point
of view mainly concerned with (1) inefficiancy of the Service
Support unit of the company and (2) insufficient quality of the
information provided. From a cognitive or mental point of view,
their stress lies on the interface between the information pro-
vider and the user.
Thus far the division of the subjects into specific groups
has given an indication of a cultural way of seeing reality or
constraining the phenomenon of information management. Based on
this indication, we will now proceed with this kind of analysis
and give a structural description of the mental processes as they
emerge from the analysis of the German text. Table 4 presents the
t-values and confidence intervals for the significant groupings.
As can be read from Table 4, five terminal states are significant-
ly different in the Figure component. There is also enough signi-
ficant differentiation in the perspective components to extract
at least one focal point. The structural information is graphic-
ally presented in Figure 3. while the English workers are con-
cerned with Organization of Mediation, their German counterpart
shows an initial interest in the Acquisition of Instruction. The
function of Instruction refers to a description of the condi-
tions for learning, whereas Acquisition addresses the phase in
the learning process that lies between the apprehension of in-
formation and the performance of a task. This is also the ini-
tial state of the process. From here the process transits through
Service Key and produces the singularity High Quality Service.
One may design various learning conditions, the one ad-
dressed here requires a steering component (Key) that places the
worker into the position of judging on his own his performances.
A performance in itself does not make it possible to conclude
that learning has occurred. Therefore, the key is necessary to
show that there has been a change in performance. Consequently,
High Quality Service point towards the worker's internal capa-
bility of accompli shement , that is, completing a service task
At the next step, the process transits through the state
Table 4 . Levels of significance for Agent, Figure, and Ground
components in German material
Predicted Realized t =V~C x a and Percentiles
Clusters ( ) Deviates (a) df of the t-distributions
Agent N = 34, n = 74, m = .065, s = .136
2 5 ' 47 1 t .95 < 7 ' 74 < t .975
Figure N = 74, n = 34, m =.055 s = .137
2 7.75 1 t 95 < 10.96 < t 975
3 1.23 2 t ' #90 < 2.13 < t' 95
4 1.18 3 t>95 < 1.88 < t' 95
5 .84 4 t' 9Q < 1.88 < t[ 95
Agent N = 10, n = 15, m = .200, s = .250
2 2.67 1 t >9Q < 3.78 < t >95
Ground N = 15, n = 10, m = .171, s = .205
7 3 -° 6 1 *.90 < 4 - 32 < *.9S
\ M High Quality Service
\ Service Degradation
\ Competence Reduction
\ Knowledge Maintenance
Figure 3. Operating structural relations
characterizing perspectives and
viewpoints of West German mecha-
named Lack of Explanation. When the worker is called upon to
show what he has learnt or can put to use in some way, he must
recognize the problem. This condition requires not only that he
can differentiate between a variety of technical details and
events which he has become familiar with in different contexts.
What he needs above all is a conceptual, that is, an explanato-
ry anchorage in order to solve some " odd " problem. Lack of
Explanation gives expression to a need for a high level instruc-
tional language. The absence of such a language results in Ser-
vice Degradation, meaning insufficiency in performance.
The next state is Information Lag, which transforms Service
Degradation into Competence Reduction. Information Lag may be
defined as a delayed feedback that produces an insufficient al-
beit negative change in mastery. To counteract this trend in cog-
nitive development, the terminal state of Control of Information
Processing transforms the process into its final state, Knowl-
edge Maintenance. The demand for control implies a function of
direction that seems to be of considerable importance to the
learning process. What is asked for is a description of the men-
tal performance expected. Knowledge Maintenance refers to the
great value of preservation of concept and conceptual relations
as a means for thinking and communication. The importance of this
state cannot be overemphasized, but since concepts are verbally
communicated, there is often a danger of loosing sight of their
physical reality. Maintenance refers to the workers' need to
know the real meaning of the concepts.
The emphasis on Knowledge Maintenance is further enhanced in
the perspective on the Figure component.
The Ground contains the reason for the German conduct. It is
defined by two terminal states, Functional Dissemination and Pro-
cessing Efficiency. The first one refers to the importance of
the dissemination of information to individual workers. The se-
cond one addresses the transformation of information into intel-
lectual capability. The result is the singularity of Measure.
Measure refers to a standard for comparison, which is necessary
in determining hypothesized progression. It is important because
the Acquisition of Instruction by necessity is based on con-
cepts and conceptual relations which require a means for ana-
lyzing the degree to which concepts are correctly referenced.
The perspective on a Figure or Ground picks up some ab-
straction, that is, some central aspects are focussed upon. It
is of special interest here that the perspective structure of
the Ground has been realizable. This gives additional weight to
the importance of Measure in information processing.
Conclusions . The consciousness of the German mechanics
points towards the problem of how information may be structured
and represented in such a way that it fosters their development
knowledgewise. The maintenance of knowledge is not only in the
focus of their perspective but interrelated with a view on ade-
quate measures controlling differentiation and structuring pro-
cesses. For them the problem of knowing is not merely a question
of permanent storing and retrieval , but rather seems to be a
highly dynamic phenomenon. Thus knowing something cannot be con-
sidered a permanent attribute of the worker, which exists during
his lifetime. It is restrained to a more or less extended time
segment of his lifetime.
That the development of competence is highly dependent on
the possibility to maintain and preserve knowledge either of the
individual or of the company is clear from the last analysis. At-
tention should therefore be paid to the question of how training
can be integrated into information processing. The text of the
Italian mechanics contains some hints that an answer might be
found there. The results of the numerical analysis show eleven
significant clusterings in the Figure component. These are given
with the clusterings of the Ground and the perspectives in Table 5.
The edges of the cubic space show ten terminal states defining the
Figure Component (Figure 4). The first terminal state concerns
Modes of Instruction. The information environment consists of va-
rious communication media, which have to be arranged in such a way
that they facilitate the performance of the workers. The specific
arrangements at a particular working place comprise the Modes of
The cognitive process starts from the state of Learning
Conditions. This state addresses a set of circumstances that
obtain when learning occurs. In a broad sense, learning refers
to certain observable changes in performance. This state is
transformed when the process transits through Instruction and
reaches Education. One might think of Instruction as a set of
procedures designed to bring out transferability or inference.
The aim is to integrate new information into previously learnt
concepts and to make the concepts distinguishable from each
Table 5 . Levels of significance for Agent, Figure, and Ground
components in Italian material
Predicted Realized t ="/c x a and Percentiles
Clusters(C) Deviates(a) df of the t-distributions
Agent N = 20, n =
40, m =
111 , s
1 . 29 2
Figure N = 40, n
= 20, m
Agent N ■ 12, n =
18, m =
Ground N = 18, n ■■
= 12, m
Utilization of Information
Utilization of Information
/ Ease of Learning
\ Intellectual Skills
Speed of Information
Sufficiency of Information
Figure 4 . Operating structural relations
characterizing perspectives and
viewpoints of Italian mechanics,
other. Specific for Education is the requirement of a learning
structure for the concepts or subjects to be acquired. Further,
Education means a specification and ordering of the prerequi-
site capabilities within a certain topic.
The next following state is Testing, whose aim is to assess
the outcome of an educational process. Testing in this sense can
either mean assessing norm-referenced or criterion-referenced.
The transition of the process through this state transforms Edu-
cation into Diagnosis. Diagnosing the educational outcome norm-
referenced implies that a distribution of scores is obtained from
entire groups. On the other hand, criterion-referenced diagnosing
means that the capability of a particular individual in attaining
defined instructional objectives is assessed. Information Provi-
sion is the next state and requires that information is communi-
cated in such a way that it can serve the particular purpose of
instruction. The transformation brought out when the process
transits this state produces the concept of Instructional Sequence.
If concepts are being learnt, various kinds of information has to
be provided in a well planned and orderly manner to gain the at-
tention of the worker. The cognitive process now transits through
Speed of Information, which allows the Instructional Sequence to
be transformed into Apprehension. The Speed of Information con-
trols the workers' possibility to attend to news and novelties.
If it is timed properly, it will facilitate the perception. Al-
though this " perceptual learning " is a necessity, Apprehension
is required if what is perceived shall be conceptualized. Once
Apprehension has occurred, the process leads over to the state
of Sufficiency of Information, which transforms it into Acquisi-
tion. Before this specific event can occur, it is necessary
to provide the workers with information that is both relevant to
their task as well as concise and precise. The phase of Acquisi-
tion is terminated as soon as one can observe that a particular
individual can execute the performance required. This state termi-
nates the first cycle in the graph and initiates the second very
From here the process leaps into Modes of Instruction, which
transforms Acquisition into Intellectual Skills. It is of im-
port to realize here that an individual worker who has been
called upon to exhibit a performance is placed in the position
not simply to tell something but rather to do something. Thus
Intellectual Skills require the instantiation of symbolic acti-
vities necessary for the performance.
The process goes on and leaps into its third and final cy-
cle, which starts with the terminal state of Tutorage. By this
is meant an interchange between a " student " and his " tutor " .
Important for this interchange is the reading by the " student "
that has been done before the meeting and that which will fol-
low after the meeting. Thus the function of the Tutorage is to
guide the " student's " thinking and to recommend the direction
of further learning. This state is transformed when the process
transits through Advantage into Ease of Learning. Strictly speak-
ing, the Advantage lies in a management function and in the as-
sessment of the "student's " performance. Ease of Learning ad-
dresses the questions the " student " can put to the " tutor " ,
because the answers provide the judgmental ground for engagement
in self-directed study.
The third cycle now crosses the second point of intersection
giving rise to the highest point in the graph, named Enablement.
The circle around the node indicates that this concept has a
higher depth in the mental structure compared to all the others.
Enablement marks the circumstance that cognitive abilities func-
tion as mediators which enable the individual to interact with
or interpret fairly discrete environmental experiences. Develop-
ing stable abilities and comprehensive knowledge requires produc-
tive learning experiences over an extended period of time. Thus
the consciousness of the Italian mechanics can concisely be ex-
pressed in their willingness to extend their abilities.
The perspective on the Figure is a little more elaborated
compared to the previous ones. In this context, the perspective
transformation is not merely a displacement of viewpoints or a
shift of the visual field, but a restructuring. Therefore, the
graphical representation of the perspective gives expression to
a new structural connection, which focusses upon Intellectual
Skills and Ease of Learning. These are lifted up from the Fi-
gure by the following routine. From the cluster analysis out-
put the typology of the Agents is determined. The typology shows
that Acquisition is the category name describing the Agent group
that defines the initial state where the transformation process
starts. The next Agent group is named Modes of Instruction. The
highest point, that is, the singularity of the curve that con-
nects both states is Intellectual Skills, which is extracted.
The final significant Agent grouping is named Tutorage. The
highest point between the singularity extracted and the last
state defining the edges of the perspective is Ease of Learning.
One would perhaps expect that Enablement would be the focal
point of the perspective. Why is this not the case? From the pre-
vious discussion of the Figure component it seems quite obvious
that the Italian workers express an interest in rising their lev-
el of proficiency. Abilities, on the other hand, are more gen-
eral and inclusive than Skills. It is the acquiring of Intellec-
tual Skills that matches Ease of Learning.
The Ground component is defined by two terminal states, In-
formation Personnel and Utilization of Information. The first
state refers to a need for persons who can handle growing quanti-
ties of information. When the process transits through the se-
cond state, a transformation takes place that results in Informa-
tion Structuring. This Structuring has to be accomplished on
terms of the workers.
The point reached is further enhanced in that it is also
the focal point of the perspective on the Ground.
Conclusions . Enablement is the root of the mental struc-
ture depicted in Figure 4. This concept stresses a quest for cog-
nitive power. The central theme of this development is that abi-
lities can be learnt. Consequently, it is of great importance
to the Italian workers that efforts be concentrated on the devel-
opment of their abilities, rather than putting efforts on striv-
ing for mastery of content in the instructional material, that
is, for factual knowledge.
The analysis of the verbal data derived from the text pro-
duced by the US mechanics has led to eleven clusterings. The
numerical values for these clusterings are given in Table 6.
The Table also shows two significant clusterings of the Agents
related to the Figure component, while those related to the
Ground component show four clusterings. The Agents of the Means
component show no significant clustering above the chosen lower
bound of t Qn of the confidence interval for the t-distributions.
The Means component itself shows two significant groupings. The
realized collinear groupings of the US material is graphically
presented in Figure 5. The discussion of the operating structur-
al relations in the US material precedes in the same way as be-
fore. The only new information to be added for the understanding
of the cube relates to the Means component. It is represented in
the upper left part of the left hand side.
The cognitive process depicted in the background starts in
the terminal state of Declination. The import of this state is
that the amount of information delivered has been exposed to va-
rious forms of destructive tendencies. For example, the steering
and controlling mechanism is conceived to be producing too much
redundancy or failure in addressing the individual worker or in
regulating the information fed back by the worker. Then Declina-
tion follows, which implies an information system whose functions
are too clumsy to make possible a differentiated growth of infor-
Personalization is the next state through which the process
transits. The indication here is that humans like to be informed
by humans. This type of information is perceived as being of con-
siderable help. So this state transforms Declination into Enable-
ment. From here the process leaps into Coordination, meaning that
a particular person functions as a selector and instructor. If
information is structured in such a way that interpersonal rela-
tions become used, it is conceived as the best information one
can get. Especially novel information requires that the worker
gets clear-cut instructions on how to treat it. The process now
develops into Constructiveness, a concept expressing a behav-
iour that involves the utilization of materials and tools in
serving to be helpful. The state of Information Access trans-
forms Constructiveness into Cognitive Operation. This is the
ultimate mood for the conscious reformation of one's own abili-
ties. Reasonableness directs attention to the consciousness
Table 6 . Levels of significannce for Agent, Figure, and Ground
components in US material
t =Vc x
a and Percentiles
Deviates (a) df
of the t
Agent N = 32, n =
< t .975
Figure N = 112, n =
= 32, m
s = .175
< t .975
< t .99
< t .995
< t .995
< t 1.00
< t .995
< t .975
< t .975
< t .975
< t .95
Agent N 13, n = 36
, m as .
, s =
< t .95
Ground N = 36, n =
— * m
< t .975
< t .975
1 . 00 3
< t .95
Means N = 9, n = 6
, m = .
, s =
< t .975
Coordination! «... .'
~ * '• * Worthiness \
Constructiveness ». _. .... 7.
» % Recognition *~
'*-,•* of Ability
Reasonableness ><'' V.^Mastery >| Lack of information
v ,-— «* Framing
Figure 5 . Operating structural relations
characterizing perspectives and
viewpoints of US mechanics. From
" A cognitive economics approach
to information management " by
B. Bierschenk, I. Bierschenk,
forskning (13), P 5.
developed by the workers and the mental activities they have
experienced. When the process transits through this state, it
develops into Recognition of Ability. This singularity denotes
a desire to pay attention to the consequences of the way in
which information is disseminated. As a consequence, devices
have to be devloped for an unbiased assessment of the workers'
abilities. Otherwise misunderstanding and misinterpretation may
result and render the company incoherent or even incomprehen-
sible. However, through Methodological Innovation, this state
can be transformed into Mastery. On the other hand, the workers
seem to show estrangement because of Information Lag, which
transforms Mastery into an insufficient Authorization. Thus at
the highest point of the first cycle in the graph, the workers
give expression to an interference with their possible achieve-
ment, which is an unsatisfactory state from their point of view.
From here the process continues into the first state of the
second cycle, namely Lack of Information. An overabundance of in-
formation without any novelty tends to be conceived as missing.
At the next step, the process reaches Information Timing, a state
that may be regarded as a positive and compensatory regulation
of this " system error " , resulting in Provision. This is also
the peak of the second cycle. This cycle crosses the path of the
first one with the effect of a deepening of the mental structure,
indicated by a circle around the node named Edification. This
concept may be conceived as a necessary means for increasing the
subjective consciousness of the workers to facilitate an inner-
directed change of their behaviour, instead of an outer-directed
(speak company-directed) by behavioural modification. The process
finally transits through the state of Pretension, which is deter-
mined by the ratio of actual behaviour to supposed behavioural
potential. The transformation taking place results in the hihgest
point of the entire curve, that is, worthiness. Thus a conception
of merit and respect has emerged as the root of the mental struc-
ture of the US mechanics.
This point is lifted up and focussed upon in the perspective
on the Figure.
The transformational process depicted in the Ground is more
elaborated than in the previous Grounds. The process has one
cycle, which shows a thematic development on efficient informa-
tion management as foundation for competence development. The
states defining the edges of the Ground, Facilitation of Infor-
mation and Concentration, indicate that infrastructures may de-
velop that require a certain Framing of information. Their Loca-
tion implies that information is organized in a manner which ap-
pears alien to the workers. Only by developing integrating and
interacting mechanisms a Focussing on required information will
be possible. For these workers the only way out of this
state is provided through the function of Tutorage, which trans-
forms Focussing into Comprehension.
The perspective on the Ground lifts up and accentuates the
highest point of this curve, namely Comprehension.
The Means component, finally, shows Information Use as the
initial state. It indicates the importance of information in the
comprehension process. The state of Supply Function transforms
Information Use into Regulation, thus expressing a need for put-
ting values on the presentation of information.
Conclusions . The textual transformations abstracted from
the US material give expression to a willingness in solving ser-
vice problems. On the other hand, the workers indicate that their
ability is insufficiently recognized. Moreover, they bring out
a self-consciousness in that service information is of concern
for them. It is considered a necessity and as such an integrative
part of their everyday life. They relate the challenge they expe-
rience directly to their Worthiness.
In this article we have tried to show how coordinative struc-
tures are modulated by information coming from various cultural
sources. We have tried to show that there exist subtle and mutu-
ally dependent relationships between textual Agents and their Ob-
jectives. Both Agent and Objectives provide structural informa-
tion relevant to the linkage as well as to textual dynamics. The
article presents five analyses which show that the grammar of Ian-
guage does not uniquely determine the development of the theme.
Each one introduces a novelty. Some of the characteristics that
emerge are noteworthy.
The concept of consciousness, especially self-consciousness,
has been studied by Sperry (1983). He defines consciousness as
the emergence of a unity of mental experience in humans. This
unity gets its expression in language with the help of an " I "
or what is substituted for it in the verbal expression of alter-
native consequences of an action (jaynes, 1976). On this basis
it can be stated that the study of consciousness focusses upon
the problems of .everyday life. For the Swedish workers, this
kind of problem is basically economic. They relate information
use to their working load within the context of ergonomic condi-
tions. The English workers seem to be conscious of the import of
the mediation of information. For them, the quality of informa-
tion is dependent on forms of distribution. Connected with it is
their conception of coherence, which implies that attention is
paid to the way in which various functions of the service support
system interacts with their working environment. The conscious-
ness of the West German workers demonstrates that they know how
to acquire instructions. They can differentiate between correct
and erroneous service by applying their personal system of con-
cepts. But they are also conscious of the fugitive nature of
knowledge, which calls for controls and measures in maintaining
it. The question that concerns the Italian workers the most is
mainly procedural in kind. For them, it is of import how a per-
son's ability can be developed to discriminate between the right
and the wrong alternatives in service tasks. It is of less im-
port for them to structure information by their own. On the con-
trary, they ask for persons who can demonstrate how to chose the
correct alternative. The consciousness of the US workers centers
around the individual's responsibility at his working place. Eve-
ry person is seen to have the responsibility for his actions,
that is, becoming informed or having access to the latest devel-
opments within his field. This strong conviction puts a heavy
stress on worthiness. The moral point of view places the US
worker in an advantage position, especially when one considers
his desire for comprehension compared to persons who have been
forced to give up responsibility or have not yet adopted it.
The US workers' understanding of what has to be done is clear
Bierschenk, B. (1982). An ecological model for the processing of
symbolic information. Perceptual and Motor Skills , 54 , 663-674.
Bierschenk, B. (1984). Steering mechanisms for knowability. Kog-
nitionsvetenskaplig forskning (1).(ERIC No. ED 264 246)
Bierschenk, B.,& Bierschenk, I. (1976). A system for a computer-
based content analysis of interview data . Lund.* Gleerup.
Bierschenk, B.,& Bierschenk, I. (1986 a). Concept formulation.
Part I. The phenomenon of cognition. Kognitionsvetenskaplig
Bierschenk, B.,& Bierschenk, I. (1986 b). Concept formulation,
part II. Mesurement of formulation processes. Kognitionsveten-
skaplig forskning (11).
Bierschenk, B., & Bierschenk, I. (1986 c). Concept formulation.
Part III. Analysis of mentality. Kognitionsvetenskaplig forsk-
Bierschenk, B. ? & Bierschenk, I. (1986 d). A cognitive economics
approach to information management. Kognitionsvetenskaplig
forskning ( 13) .
De Mey, M. (1982). The cognitive paradigm. Cognitive science, a
newly explored approach to the study of cognition applied in
the analysis of science and scientific knowledge . Dordrecht :
Jaynes, J. (1976). The origin of consciousness in the breakdown
of the bicameral mind . Boston: Houghton Mifflin.
Pribram, K. (1986). The cognitive revolution and mind/brain issues,
American Psychologist , 41 . 507-520.
Sperry, R. (1983). Science and morality priority . Oxford: Basil
Ward, J.H. (1963). Hierarchical grouping to optimize an objec-
tive function. Journal of American Statistical Association ,
Wigner, E.P. (1969). Epistemology of quantum mechanics: Its ap-
praisals and demands. In M. Grene (Ed.), The anatomy of knowl -
edge (pp 31-45) .
Wishart, D. (1982). Clustan: User manual . ( Inter-University Re-
search Council. Series No. 47). Edinburgh: Edinburgh Univer-
sity, Program Library Unit.
This research was supported by a grant from Volvo Car Cor-
poration, Goteborg, Sweden. The open-ended questions were formu-
lated by their Service Support unit, which also collected the
Parts of this article were presented at the 10th Scandina-
vian Conference of Business Administration, August 19-22 1986
at the Norwegian School of Economics and Business Administration
in Bergen, Norway.